Research on Magnetic Resonance Imaging of human coronary arteries is being carried out under a current contract from the NHLBI. We have demonstrated that MRI detects about half of the lesions found by coronary arteriography in the left main, left anterior descending, proximal circumflex, and proximal right coronary arteries. Key ingredients in achieving these results are the use heart-synchronized gating and of cycled multi-section imaging. The hardware used in this study continues to improve in performance and this offers new opportunities for technical improvements that may results in better clinical results. Also, different techniques can be explored. Specifically, we propose to investigate 1. Optimization of in-plane spatial resolution to match the needs of coronary images; 2. Investigation of the effects of section thickness on vessel conspicuity, including post-processing by selective addition of sections; 3. Increasing time resolution, to sample the position of the arteries at 50 msec rather than 100 msec intervals; 4. Angling of the imaging plane, both during acquisition and by post-processing to improve vessel visualization; 5. Using partial flip imaging to decrease signal from the myocardium and increase the signal from blood, so as to make the arteries show as positive contrast features. This component will include an analytical approach to developing the optimal sequence parameters; 6. use of phase information to detect flow; 7. Development of software to provide three-dimensional maps of the coronary vasculature; and 8. Characterization of echo planar imaging of the human coronary arteries. Technical validations will be done with normal volunteers, and clinical evaluations will be performed in patients who have undergone coronary arteriography. Because of its non-invasiveness, MRI is ideal for longitudinal studies designed to lead to an understanding of human atherosclerotic disease.